Demonstration of fusion material calculations using quantum computers — ORNL, US Medical Institution, and IBM mark world first

IBM announced on July 6 that a joint research team from Oak Ridge National Laboratory (ORNL), Cleveland Clinic, and IBM successfully calculated nine molecular structures of a material promising for fusion energy fuel production. While this is a computational challenge that is difficult to handle at scale using classical computers alone, the research team succeeded by integrating quantum computers with classical systems. This is said to be the world's first instance of performing such calculations using quantum computers.

Securing a sufficient amount of tritium has been a long-standing challenge toward realizing clean and abundant energy through fusion power plants. Solving this challenge is one of the key goals of the United States Department of Energy (DOE) Genesis Mission.

Quantum computers are suitable for atomic-level chemical calculations of a liquid salt containing fluorine, lithium, and beryllium (FLiBe). Among such materials, FLiBe is a leading candidate for extracting tritium fuel in fusion reactors.

To calculate various structures of FLiBe clusters, the research team utilized quantum-centric supercomputing technology, which is also used in a 12,635-atom scale protein simulation currently being conducted with Cleveland Clinic. These methods calculate the quantum behavior of electrons in complex materials, complementing and enhancing the capabilities of classical supercomputers and algorithms.

Tom Beck, Section Head for Science Engagement in the Computing and Computational Sciences Directorate at ORNL, said: "In order to demonstrate the capabilities catalyzed by the Genesis Mission, we have built a team of leading experts across seven DOE national labs, four universities, three industry partners, and Cleveland Clinic to pursue a multi-pronged discovery cycle aimed at optimizing tritium production in molten salt fusion blanket materials. Quantum computers, such as those built by IBM and enhanced by AI and exascale computing, are key tools that accelerate the discovery and design cycles needed to produce sufficient tritium to fuel fusion reactors."

As one of the industrial partners of the mission, IBM is exploring with its partners the possibilities of quantum-centric supercomputing that combines CPUs, GPUs, and QPUs to tackle problems that are difficult to solve using only one type of the aforementioned processors alone.

Finding the optimal composition of FLiBe, whose structure constantly changes under the influence of strong neutron irradiation, extreme temperatures, and magnetic fields, is one of the most challenging tasks in current science and engineering. To achieve this, it is necessary not only to study quantum mechanical properties such as energy characteristics, stability, and interactions with tritium in detail, but also to understand how it performs multiple roles, including its function as a tritium breeding material in high-temperature environments.

Currently, such research can only be conducted through difficult and costly experiments, or through approximation calculations using classical computing with limited precision.

Therefore, the research team utilized quantum-centric supercomputing to link quantum computers with classical computers in order to calculate the energies of various FLiBe structures with and without tritium.

In this approach, the parts of the problem that can be expressed as quantum circuits are processed by quantum computers. This made it possible to reveal, with higher accuracy, the electronic structure of the material and atomic behaviors, particularly how strongly it binds with tritium at the molecular level.

Furthermore, the team identified the range of structures in which atoms change, revealing previously unobtainable characteristics such as the mechanism and strength with which each structure binds to tritium.

This collaboration is ongoing, and the research team is working to reduce data transfer times between quantum and classical computing environments and expand the scale of molecular interactions that can be simulated.